Date: Fri, 11 Oct 1996 02:00:02 -0700 (PDT) From: Jake Hamby <jehamby@lightside.com> To: hackers@freebsd.org Subject: Whoohoo! JDK works! (was Re: JDK 1.0.2 SIGSEGV?!) Message-ID: <Pine.AUX.3.94.961011015108.12540A-100000@covina.lightside.com> In-Reply-To: <Pine.AUX.3.94.961005210335.25920B-100000@covina.lightside.com>
next in thread | previous in thread | raw e-mail | index | archive | help
When I complain about something, I usually like to follow up if I manage to fix the problem. In this case (the SIGSEGV error I was complaining about with "unsupported" FreeBSD JDK 1.0.2 on freefall), I'm proud to say the problem is solved. Obviously, something changed in -current between the time the JDK 1.0.2 was built, and when I ran it. Something that could crash the JDK very early on, before it even printed a startup message. Libc would be a candidate, as would libm. But let's think, what major component of our shared library architecture has changed recently (even prompting a major outcry as the original patch caused machines to crash)? That's right: ld.so. Call it intuition, but I suspected ld.so in the back of my mind all along, but didn't feel like grabbing an older copy out of CVS to play with. Long story short, when I CVSup'ed -current tonight and saw rtld.c had changed (twice), I decided to rebuild ld.so, install it, cross my fingers, say a little prayer, start java, expecting the worst, and ... it worked! Let that be a lesson to all FreeBSD hackers (actually two lessons): 1) First impressions are almost always correct, and 2) There is heavy magic in ld.so! I can't wait to play with Java tomorrow, on my favorite Free UNIX! Whoohoo... -- Jake
Want to link to this message? Use this URL: <https://mail-archive.FreeBSD.org/cgi/mid.cgi?Pine.AUX.3.94.961011015108.12540A-100000>